I came across the source code in github about this feature, which I believe is very useful to compete against other API Gateway vendors. But it was not included in the main bench. Are does anyone know why and if there is any plan to add it in the future?
The increment happens in the log phase, asynchronously:
So the counters are incremented async, but we still get usage synchronously for each call and that incurs a round trip call to the data store/redis, right?
To further reduce the latency, can I suggest a feature to retrieve the counter periodically such as based on some time interval or based on the number of accumulated calls?
Do we have any rough idea about the time for using Cassandra vs Redis?
Thanks for your help!
This is available in the enterprise version of the plugin:
Secure, Manage & Extend your APIs or Microservices with plugins for authentication, logging, rate-limiting, transformations and more.
It’s difficult to have an accurate number here as the details matter a lot here but you should see sub-milisecond response times from Redis ususally if you have Redis running in the same network.